zfs/include/sys/abd.h

189 lines
5.0 KiB
C
Raw Normal View History

/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2014 by Chunwei Chen. All rights reserved.
single-chunk scatter ABDs can be treated as linear Scatter ABD's are allocated from a number of pages. In contrast to linear ABD's, these pages are disjoint in the kernel's virtual address space, so they can't be accessed as a contiguous buffer. Therefore routines that need a linear buffer (e.g. abd_borrow_buf() and friends) must allocate a separate linear buffer (with zio_buf_alloc()), and copy the contents of the pages to/from the linear buffer. This can have a measurable performance overhead on some workloads. https://github.com/zfsonlinux/zfs/commit/87c25d567fb7969b44c7d8af63990e ("abd_alloc should use scatter for >1K allocations") increased the use of scatter ABD's, specifically switching 1.5K through 4K (inclusive) buffers from linear to scatter. For workloads that access blocks whose compressed sizes are in this range, that commit introduced an additional copy into the read code path. For example, the sequential_reads_arc_cached tests in the test suite were reduced by around 5% (this is doing reads of 8K-logical blocks, compressed to 3K, which are cached in the ARC). This commit treats single-chunk scattered buffers as linear buffers, because they are contiguous in the kernel's virtual address space. All single-page (4K) ABD's can be represented this way. Some multi-page ABD's can also be represented this way, if we were able to allocate a single "chunk" (higher-order "page" which represents a power-of-2 series of physically-contiguous pages). This is often the case for 2-page (8K) ABD's. Representing a single-entry scatter ABD as a linear ABD has the performance advantage of avoiding the copy (and allocation) in abd_borrow_buf_copy / abd_return_buf_copy. A performance increase of around 5% has been observed for ARC-cached reads (of small blocks which can take advantage of this), fixing the regression introduced by 87c25d567. Note that this optimization is only possible because all physical memory is always mapped into the kernel's address space. This is not the case for HIGHMEM pages, so the optimization can not be made on 32-bit systems. Reviewed-by: Chunwei Chen <tuxoko@gmail.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Closes #8580
2019-06-11 16:02:31 +00:00
* Copyright (c) 2016, 2019 by Delphix. All rights reserved.
*/
#ifndef _ABD_H
#define _ABD_H
#include <sys/isa_defs.h>
#include <sys/debug.h>
#include <sys/refcount.h>
#include <sys/uio.h>
#ifdef __cplusplus
extern "C" {
#endif
typedef enum abd_flags {
ABD_FLAG_LINEAR = 1 << 0, /* is buffer linear (or scattered)? */
ABD_FLAG_OWNER = 1 << 1, /* does it own its data buffers? */
ABD_FLAG_META = 1 << 2, /* does this represent FS metadata? */
ABD_FLAG_MULTI_ZONE = 1 << 3, /* pages split over memory zones */
single-chunk scatter ABDs can be treated as linear Scatter ABD's are allocated from a number of pages. In contrast to linear ABD's, these pages are disjoint in the kernel's virtual address space, so they can't be accessed as a contiguous buffer. Therefore routines that need a linear buffer (e.g. abd_borrow_buf() and friends) must allocate a separate linear buffer (with zio_buf_alloc()), and copy the contents of the pages to/from the linear buffer. This can have a measurable performance overhead on some workloads. https://github.com/zfsonlinux/zfs/commit/87c25d567fb7969b44c7d8af63990e ("abd_alloc should use scatter for >1K allocations") increased the use of scatter ABD's, specifically switching 1.5K through 4K (inclusive) buffers from linear to scatter. For workloads that access blocks whose compressed sizes are in this range, that commit introduced an additional copy into the read code path. For example, the sequential_reads_arc_cached tests in the test suite were reduced by around 5% (this is doing reads of 8K-logical blocks, compressed to 3K, which are cached in the ARC). This commit treats single-chunk scattered buffers as linear buffers, because they are contiguous in the kernel's virtual address space. All single-page (4K) ABD's can be represented this way. Some multi-page ABD's can also be represented this way, if we were able to allocate a single "chunk" (higher-order "page" which represents a power-of-2 series of physically-contiguous pages). This is often the case for 2-page (8K) ABD's. Representing a single-entry scatter ABD as a linear ABD has the performance advantage of avoiding the copy (and allocation) in abd_borrow_buf_copy / abd_return_buf_copy. A performance increase of around 5% has been observed for ARC-cached reads (of small blocks which can take advantage of this), fixing the regression introduced by 87c25d567. Note that this optimization is only possible because all physical memory is always mapped into the kernel's address space. This is not the case for HIGHMEM pages, so the optimization can not be made on 32-bit systems. Reviewed-by: Chunwei Chen <tuxoko@gmail.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Closes #8580
2019-06-11 16:02:31 +00:00
ABD_FLAG_MULTI_CHUNK = 1 << 4, /* pages split over multiple chunks */
ABD_FLAG_LINEAR_PAGE = 1 << 5, /* linear but allocd from page */
} abd_flags_t;
typedef struct abd {
abd_flags_t abd_flags;
uint_t abd_size; /* excludes scattered abd_offset */
struct abd *abd_parent;
zfs_refcount_t abd_children;
union {
struct abd_scatter {
uint_t abd_offset;
#if defined(__FreeBSD__) && defined(_KERNEL)
uint_t abd_chunk_size;
void *abd_chunks[];
#else
uint_t abd_nents;
struct scatterlist *abd_sgl;
#endif
} abd_scatter;
struct abd_linear {
void *abd_buf;
single-chunk scatter ABDs can be treated as linear Scatter ABD's are allocated from a number of pages. In contrast to linear ABD's, these pages are disjoint in the kernel's virtual address space, so they can't be accessed as a contiguous buffer. Therefore routines that need a linear buffer (e.g. abd_borrow_buf() and friends) must allocate a separate linear buffer (with zio_buf_alloc()), and copy the contents of the pages to/from the linear buffer. This can have a measurable performance overhead on some workloads. https://github.com/zfsonlinux/zfs/commit/87c25d567fb7969b44c7d8af63990e ("abd_alloc should use scatter for >1K allocations") increased the use of scatter ABD's, specifically switching 1.5K through 4K (inclusive) buffers from linear to scatter. For workloads that access blocks whose compressed sizes are in this range, that commit introduced an additional copy into the read code path. For example, the sequential_reads_arc_cached tests in the test suite were reduced by around 5% (this is doing reads of 8K-logical blocks, compressed to 3K, which are cached in the ARC). This commit treats single-chunk scattered buffers as linear buffers, because they are contiguous in the kernel's virtual address space. All single-page (4K) ABD's can be represented this way. Some multi-page ABD's can also be represented this way, if we were able to allocate a single "chunk" (higher-order "page" which represents a power-of-2 series of physically-contiguous pages). This is often the case for 2-page (8K) ABD's. Representing a single-entry scatter ABD as a linear ABD has the performance advantage of avoiding the copy (and allocation) in abd_borrow_buf_copy / abd_return_buf_copy. A performance increase of around 5% has been observed for ARC-cached reads (of small blocks which can take advantage of this), fixing the regression introduced by 87c25d567. Note that this optimization is only possible because all physical memory is always mapped into the kernel's address space. This is not the case for HIGHMEM pages, so the optimization can not be made on 32-bit systems. Reviewed-by: Chunwei Chen <tuxoko@gmail.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Closes #8580
2019-06-11 16:02:31 +00:00
struct scatterlist *abd_sgl; /* for LINEAR_PAGE */
} abd_linear;
} abd_u;
} abd_t;
typedef int abd_iter_func_t(void *buf, size_t len, void *private);
typedef int abd_iter_func2_t(void *bufa, void *bufb, size_t len, void *private);
extern int zfs_abd_scatter_enabled;
static inline boolean_t
abd_is_linear(abd_t *abd)
{
return ((abd->abd_flags & ABD_FLAG_LINEAR) != 0 ? B_TRUE : B_FALSE);
}
single-chunk scatter ABDs can be treated as linear Scatter ABD's are allocated from a number of pages. In contrast to linear ABD's, these pages are disjoint in the kernel's virtual address space, so they can't be accessed as a contiguous buffer. Therefore routines that need a linear buffer (e.g. abd_borrow_buf() and friends) must allocate a separate linear buffer (with zio_buf_alloc()), and copy the contents of the pages to/from the linear buffer. This can have a measurable performance overhead on some workloads. https://github.com/zfsonlinux/zfs/commit/87c25d567fb7969b44c7d8af63990e ("abd_alloc should use scatter for >1K allocations") increased the use of scatter ABD's, specifically switching 1.5K through 4K (inclusive) buffers from linear to scatter. For workloads that access blocks whose compressed sizes are in this range, that commit introduced an additional copy into the read code path. For example, the sequential_reads_arc_cached tests in the test suite were reduced by around 5% (this is doing reads of 8K-logical blocks, compressed to 3K, which are cached in the ARC). This commit treats single-chunk scattered buffers as linear buffers, because they are contiguous in the kernel's virtual address space. All single-page (4K) ABD's can be represented this way. Some multi-page ABD's can also be represented this way, if we were able to allocate a single "chunk" (higher-order "page" which represents a power-of-2 series of physically-contiguous pages). This is often the case for 2-page (8K) ABD's. Representing a single-entry scatter ABD as a linear ABD has the performance advantage of avoiding the copy (and allocation) in abd_borrow_buf_copy / abd_return_buf_copy. A performance increase of around 5% has been observed for ARC-cached reads (of small blocks which can take advantage of this), fixing the regression introduced by 87c25d567. Note that this optimization is only possible because all physical memory is always mapped into the kernel's address space. This is not the case for HIGHMEM pages, so the optimization can not be made on 32-bit systems. Reviewed-by: Chunwei Chen <tuxoko@gmail.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Closes #8580
2019-06-11 16:02:31 +00:00
static inline boolean_t
abd_is_linear_page(abd_t *abd)
{
return ((abd->abd_flags & ABD_FLAG_LINEAR_PAGE) != 0 ?
B_TRUE : B_FALSE);
}
/*
* Allocations and deallocations
*/
abd_t *abd_alloc(size_t, boolean_t);
abd_t *abd_alloc_linear(size_t, boolean_t);
abd_t *abd_alloc_for_io(size_t, boolean_t);
abd_t *abd_alloc_sametype(abd_t *, size_t);
void abd_free(abd_t *);
abd_t *abd_get_offset(abd_t *, size_t);
abd_t *abd_get_offset_size(abd_t *, size_t, size_t);
abd_t *abd_get_from_buf(void *, size_t);
void abd_put(abd_t *);
/*
* Conversion to and from a normal buffer
*/
void *abd_to_buf(abd_t *);
void *abd_borrow_buf(abd_t *, size_t);
void *abd_borrow_buf_copy(abd_t *, size_t);
void abd_return_buf(abd_t *, void *, size_t);
void abd_return_buf_copy(abd_t *, void *, size_t);
void abd_take_ownership_of_buf(abd_t *, boolean_t);
void abd_release_ownership_of_buf(abd_t *);
/*
* ABD operations
*/
int abd_iterate_func(abd_t *, size_t, size_t, abd_iter_func_t *, void *);
int abd_iterate_func2(abd_t *, abd_t *, size_t, size_t, size_t,
abd_iter_func2_t *, void *);
void abd_copy_off(abd_t *, abd_t *, size_t, size_t, size_t);
void abd_copy_from_buf_off(abd_t *, const void *, size_t, size_t);
void abd_copy_to_buf_off(void *, abd_t *, size_t, size_t);
int abd_cmp(abd_t *, abd_t *);
int abd_cmp_buf_off(abd_t *, const void *, size_t, size_t);
void abd_zero_off(abd_t *, size_t, size_t);
Update build system and packaging Minimal changes required to integrate the SPL sources in to the ZFS repository build infrastructure and packaging. Build system and packaging: * Renamed SPL_* autoconf m4 macros to ZFS_*. * Removed redundant SPL_* autoconf m4 macros. * Updated the RPM spec files to remove SPL package dependency. * The zfs package obsoletes the spl package, and the zfs-kmod package obsoletes the spl-kmod package. * The zfs-kmod-devel* packages were updated to add compatibility symlinks under /usr/src/spl-x.y.z until all dependent packages can be updated. They will be removed in a future release. * Updated copy-builtin script for in-kernel builds. * Updated DKMS package to include the spl.ko. * Updated stale AUTHORS file to include all contributors. * Updated stale COPYRIGHT and included the SPL as an exception. * Renamed README.markdown to README.md * Renamed OPENSOLARIS.LICENSE to LICENSE. * Renamed DISCLAIMER to NOTICE. Required code changes: * Removed redundant HAVE_SPL macro. * Removed _BOOT from nvpairs since it doesn't apply for Linux. * Initial header cleanup (removal of empty headers, refactoring). * Remove SPL repository clone/build from zimport.sh. * Use of DEFINE_RATELIMIT_STATE and DEFINE_SPINLOCK removed due to build issues when forcing C99 compilation. * Replaced legacy ACCESS_ONCE with READ_ONCE. * Include needed headers for `current` and `EXPORT_SYMBOL`. Reviewed-by: Tony Hutter <hutter2@llnl.gov> Reviewed-by: Olaf Faaland <faaland1@llnl.gov> Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Pavel Zakharov <pavel.zakharov@delphix.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> TEST_ZIMPORT_SKIP="yes" Closes #7556
2018-02-16 01:53:18 +00:00
#if defined(_KERNEL)
unsigned int abd_scatter_bio_map_off(struct bio *, abd_t *, unsigned int,
size_t);
unsigned long abd_nr_pages_off(abd_t *, unsigned int, size_t);
#endif
void abd_raidz_gen_iterate(abd_t **cabds, abd_t *dabd,
ssize_t csize, ssize_t dsize, const unsigned parity,
void (*func_raidz_gen)(void **, const void *, size_t, size_t));
void abd_raidz_rec_iterate(abd_t **cabds, abd_t **tabds,
ssize_t tsize, const unsigned parity,
void (*func_raidz_rec)(void **t, const size_t tsize, void **c,
const unsigned *mul),
const unsigned *mul);
/*
* Wrappers for calls with offsets of 0
*/
static inline void
abd_copy(abd_t *dabd, abd_t *sabd, size_t size)
{
abd_copy_off(dabd, sabd, 0, 0, size);
}
static inline void
abd_copy_from_buf(abd_t *abd, const void *buf, size_t size)
{
abd_copy_from_buf_off(abd, buf, 0, size);
}
static inline void
abd_copy_to_buf(void* buf, abd_t *abd, size_t size)
{
abd_copy_to_buf_off(buf, abd, 0, size);
}
static inline int
abd_cmp_buf(abd_t *abd, const void *buf, size_t size)
{
return (abd_cmp_buf_off(abd, buf, 0, size));
}
static inline void
abd_zero(abd_t *abd, size_t size)
{
abd_zero_off(abd, 0, size);
}
/*
* Module lifecycle
*/
void abd_init(void);
void abd_fini(void);
#ifdef __cplusplus
}
#endif
#endif /* _ABD_H */